AI-DRIVEN METHODS FOR REAL-TIME MONITORING OF CROWD CONGESTION AND ABNORMAL ACTIVITIES IN PUBLIC SPACES

Abstract

The rapid growth of urban populations and the increasing use of public spaces
have intensified the need for effective crowd monitoring and security management
systems. This study investigates artificial intelligence (AI)–driven methods for realtime
monitoring of crowd congestion and detection of abnormal activities in public
spaces. The proposed approach integrates computer vision, deep learning, and
behavioral analysis techniques to process video surveillance data and extract
meaningful spatiotemporal patterns. Crowd density estimation and flow analysis are
employed to identify congestion zones, while anomaly detection models are used to
recognize unusual movement patterns and suspicious behaviors. Experimental
evaluations demonstrate that AI-based methods significantly improve detection
accuracy, responsiveness, and scalability compared to traditional surveillance
systems.